## Hauptseminar Digitale Kommunikationssysteme

Vortragende/r (Mitwirkende/r) | |
---|---|

Nummer | 820085453 |

Art | Seminar |

Umfang | 3 SWS |

Semester | Sommersemester 2020 |

Unterrichtssprache | Deutsch |

Stellung in Studienplänen | Siehe TUMonline |

Termine | Siehe TUMonline |

### Termine

### Teilnahmekriterien

### Lernziele

### Beschreibung

### Lehr- und Lernmethoden

### Links

## Organization

The preliminary meeting will take place on Thursday, April 23, 2020 at 3PM. You will receive an invitation to a zoom conference, so please make sure that you know how to use zoom (camera is not mandatory) and that you can receive e-mails sent to your TUM address (check you settings at campus.tum.de if you don't know what this is about).

Please note that only students which participate in this meeting can be considered for this seminar! The final number of participants will be limited by the number of available topics and if there are too many students, those with earlier registration will be given prioriy.

Below is a list of available topics. Please pick **four** of them and send your choice including priorities to Tasnad Kernetzky latest until Wednesday, April 22, 2020 23:59. The topics will be assigned by the first come, first served principle as well.

## Polarization-adjusted convolutional codes (Mustafa Cemil Coşkun)

The student's task is to understand a recently introduced coding scheme, namely polarization-adjusted convolutional (PAC) codes [1]. In addition to encoding and decoding schemes presented in the papers below [2,3], the students asked to study the error exponent for concatenated code ensembles [4], which is closely related to the good performance of not only concatenated polar codes but also PAC codes.

### References

- [1] arxiv.org/pdf/1908.09594.pdf [2] arxiv.org/pdf/2002.06805.pdf [3] arxiv.org/pdf/2003.08640.pdf [4] yoksis.bilkent.edu.tr/pdf/files/13637.pdf

### Prerequisites

- Information Theory, Channel Coding, Channel Codes for Iterative Decoding

## Side channel attacks on Post-Quantum Cryptographic Schemes (Georg Maringer)

The student's task in this seminar is to understand state of the art attacks on the implementations of cryptographic schemes. Those methods are then applied to recent implementations of cryptographic schemes which are considered to be secure even if a potential attacker has access to a large-scale quantum computer. As an example the timing attack on the implementation of error correcting codes is studied.

### References

### Prerequisites

- Channel Coding, Security in Communications and Data Storage, Sichere Implementierung kryptographischer Verfahren (optional)

## Low-complexity Decoding on Non-Binary LDPC Codes (Emna Ben Yacoub)

Due to their outstanding error correction capability, non-binary LDPC codes have received a tremendous research interest. However, this comes at the cost of a of high decoding complexity. The purpose of this seminar is to review reduced complexity decoding algorithms.

### References

- [1] ieeexplore.ieee.org/abstract/document/4155118 [2] ieeexplore.ieee.org/abstract/document/4288786 [3] ieeexplore.ieee.org/document/6034757 [4] arxiv.org/pdf/1906.02537.pdf

### Prerequisites

- Channel Codes for Iterative Decoding

## Channel capacity for memory with defects and partial defects. (Haider Al Kim)

The Shannon capacity theorem defines the maximum amount of information, or data capacity , which can be sent over any channel or medium. C = B log 2 ( 1 + S / N ) where. C is the channel capacity in bits per second (or maximum rate of data [1]. The authors in [2] analyze the capacity of the partial defects in memory and study how far their constructions are from the channel capacity. They consider the storage channel in which a cell can be partially stuck-at level s with probability p . This channel model is an example of a discrete memoryless memory cell which was studied by Heegard and El Gamal in [3] for memory with defects . The student's task is to summarize both achieved capacities for memory with defects [3] and partial defects [2], respectively, compare them and compare them with a memory channel that does not have any kind of defects. The student can also try to compare if additional random errors happen what is the capacity then.

### References

- [1] Brémaud P. (2017) Shannon’s Capacity Theorem. In: Discrete Probability Models and Methods. Probability Theory and Stochastic Modelling, vol 78. Springer, Cham link.springer.com/chapter/10.1007/978-3-319-43476-6_12 [2] A. Wachter-Zeh, E. Yaakobi, "Codes for Partially Stuck-at Memory Cells", IEEE Transactions on Information Theory , vol. 62, no. 2, pp. 639-654, 2016. ieeexplore.ieee.org/document/7369966 [3] C. Heegard and A. E. Gamal, "On the capacity of computer memory with defects," in IEEE Transactions on Information Theory , vol. 29, no. 5, pp. 731-739, September 1983. ieeexplore.ieee.org/document/1056723/metrics

### Prerequisites

- Channel Coding, Information Theory

## Probabilistic Shaping with Polar Codes (Tobias Prinz)

The chapacity of the AWGN channel is given as 1/2log(1+SNR) and can be achieved by a Gaussian distributed input. In practical transmission schemes, mostly ASK and QAM schemes with uniformly distributed input symbols are used. In order to operate close to the capacity, "Gaussian-like" distributions on the input symbols should be used. Using non-uniformly distributed input constellations is known as probabilistic shaping [1] . Polar codes offer two options to perform probabilistic shaping. In [1], a combination of polar codes and probabilistic amplitude shaping [2] is introduced. In [3], polar codes are used directly to generate non-uniformly distributed symbols. The student's task is to summarize both methods, find applications and variations of both methods in literature, and try to compare them in terms of performance and complexity.

### References

### Prerequisites

- Channel Codes for Iterative Decoding, Coded Modulation

## Nonlinearity Compensation on the Fiber (Yizhao Jia)

With the exponential growth of the communicated information in an optical fiber, a physical limits is appeared for the transmission rate. By this topic, students can get the limitation of the capacity of fiber and the novel methods to solve that.

### References

- [1] Partha P Mitra, Jason B Stark - Nonlinear limits to the information capacity of optical fibre communications arxiv.org/pdf/physics/0011016.pdf [2] A Survey on Fiber Nonlinearity Compensation for 400 Gb/s and Beyond Optical Communication Systems ieeexplore.ieee.org/stamp/stamp.jsp

### Prerequisites

- Optical Communication Systems, Nonlinear Fiber Optics

## Polar Codes for the Deletion Channel (Thomas Wiegart)

Polar codes are the first class of channel codes that are provably capacity achieving on binary symmetric channels. Recently, they were also considered for the deletion channel, where arbitrary transmit symbols get lost while transmission (without knowing which positions are deleted). The students task is to understand and summarize the polar coding scheme for the deletion channel from [1] (and the extended version [2]).

### References

### Prerequisites

- Channel Coding, Chanel Codes for Iterative Decoding

## Codes in the Damerau Distance for Deletion and Adjacent Transposition Correction (Lorenz Welter)

Due to the increasing demand for large-scale storage systems DNA-based storage systems are investigated nowadays as a potential candidate. Due its biological nature it can function as a high-density and long-term storage medium. However, due to mutations of the DNA different errors during the storage process may occur which needs to be handled. These error types differ from the standard substitutions. Therefore, we need error-correcting codes which are capable of correcting insertions, deletions, substitutions and transpositions. The student's task is to understand the error types deletions and adjacent transpositions. This includes the introduced Damerau metric in [1] which is an extension of the Levenshtein metric. In a next step the student needs to understand how to construct codes capable of correcting these. [1]

### References

- [1] Gabrys. Yaakobi, Milenkovic - Codes in the Damerau Distance for Deletion and Adjacent Transposition Correction ieeexplore.ieee.org/abstract/document/8122061 [2] Sloane - On Single-Deletion-Correcting Codes arxiv.org/abs/math/0207197

### Prerequisites

- Channel Coding, Coding Theory for Storage and Networks (optional)

## Gradient coding for straggler mitigation in distributed computing (Marvin Xhemrishi)

With the growth of machine learning applications and in the era of big data, the interest towards distributed computing has risen. However, with the splitting of a highly complex computation into lesser ones and computation of those in parallel from external worker nodes, a lot of issues have been encountered. One of them is the presence of straggler nodes, i.e., worker nodes which are slower than the other ones. Therefore, techniques of straggler mitigation have been studied recently. One approach is gradient coding, which is used also in distributed learning. The tasks of the student is to understand how gradient coding helps in straggler mitigation [1] and how the coding theory tools can help in designing gradient codes [2].

### References

- [1] Tandon, Rashish et al. - Gradient Coding: Avoiding Stragglers in Synchronous Gradient Descent arxiv.org/pdf/1612.03301.pdf [2] Raviv, Netanel et al. - Gradient Coding from Cyclic MDS Codes and Expander Graphs arxiv.org/pdf/1707.03858.pdf

### Prerequisites

- Channel coding, Basic knowledge in algebra

## Amplifier scheme adaptive de-/normalization for NFT-based Optical Transmission Systems (Benedikt Leible)

With the ever increasing demand for higher data rates for optical fiber systems and the imminent 'capacity crunch' for wave division multiplexing (WDM) systems [4], modulation schemes aided by the nonlinear Fourier transform (NFT) have received a widespread interest in the last couple of years. Since the mathematical model underlying the NFT in many cases is far from real system specifications, effects such as noise and fiber loss degrade the transmission capabilities of such systems. One way of mitigating the negative effects of fiber loss on such systems is using path loss averaged (PLA) pulses as proposed in [1] and [2] for lumped (EDFA) and distributed (Raman) amplification respectively. The students task is to first get familiar with the basics of NFT-based systems by having a look at [3] and refresh their knowledge on optical amplification schemes. In the next step the methods used for the PLA approaches for EDFA [1] and Raman [2] amplification should be reviewed. (Please note that it is not necessary to fully understand every aspect of the NFT from a mathematical/algorithmic standpoint since this would exceed the scope of the seminar. However a basic idea of how NFT-based systems work should be acquired over the course of the preparations and a very short introduction on the concept should be given in the paper/presentation. Note also that reference [4] gives some insight in the problematic regarding the capacity crunch but is not meant to be fully reviewed. The main focus of this seminar topic is on references [1] and [2].)

### References

- [1] Le, Prilepsky, Turitsyn - Nonlinear Inverse Synthesis Technique for Optical Links with Lumped Amplification www.osapublishing.org/oe/abstract.cfm [2] Le et al. - Nonlinear Inverse Synthesis for Optical Links with Distributed Raman Amplification ieeexplore.ieee.org/abstract/document/7364155 [3] Yousefi, Kschischang - Information Transmission Using the Nonlinear Fourier Transform, Part 1: Mathematical Tools ieeexplore.ieee.org/abstract/document/6808480 [4] Winzer - Optical Networking Beyond WDM ieeexplore.ieee.org/abstract/document/6239631

### Prerequisites

- Optical Communication Systems, Nonlinear Fiber Optics

## Covert Communication with Sequential Change-Point Detection (Diego Lentner)

Covert Communication refers to a problem where one user Alice wants to communicate with a legitimate second user Bob while trying to hide their communication from a third user, warden Willie. Unlike in the information-theoretic secrecy problem, Willie is interested in detecting the presence of the communication itself rather than the content of a message. Covert communication has been shown to follow a square root law for many classes of channels [1]-[3] , i.e., the number of information bits which can be covertly transmitted increases only with the square root of the number of channel uses. Only recently, sequential change-point detection has been proposed as new framework for covert communication [4]. The task of the student will be to understand the fundamentals of covert communication and to review the sequential change-point detection approach proposed in [4].

### References

- [1] B. A. Bash, D. Goeckel, and D. Towsley, “Limits of reliable communication with low probability of detection on AWGN channels,” IEEE J. Sel. Areas Commun., vol. 31, no. 9, pp. 1921–1930, Sep. 2013.Available: ieeexplore.ieee.org/document/6584948 [2] M. R. Bloch, “Covert communication over noisy channels: A resolvability perspective,” IEEE Trans. Inf. Theory, vol. 62, no. 5, pp. 2334–2354, May 2016. Available: ieeexplore.ieee.org/document/7407378 [3] L. Wang, G. W. Wornell, and L. Zheng, “Fundamental limits of communication with low probability of detection,” IEEE Trans. Inf. Theory, vol. 62, no. 6, pp. 3493–3503, Jun. 2016. Available: ieeexplore.ieee.org/document/7447769 [4] K. Huang, H. Wang, D. Towsley and H. V. Poor, "LPD Communication: A Sequential Change-Point Detection Perspective," accepted for publication in IEEE Transactions on Communications. Available: ieeexplore.ieee.org/document/8970415

### Prerequisites

- Information Theory, Basic Knowledge in Probability Theory and Statistics

## Deep Learning for Fiber-Optical Communications (Daniel Plabst)

Intensity-Modulation and Direct-Detection (IMDD) transceivers can be built with low hardware complexity, low power consumption and small form factors, which makes them a promising approach for fiber-optical transmission over short distances. However, by their reduced hardware complexity, nonlinear distortions are introduced, which need to be compensated. The task of the student is to understand how the authors in [1] used end-to-end deep learning to communicate over an IMDD link.

### References

- [1] Boris Karanov, Mathieu Chagnon, Félix Thouin, Tobias A. Eriksson, Henning Bülow, Domaniç Lavery, Polina Bayvel, and Laurent Schmalen, "End-to-End Deep Learning of Optical Fiber Communications," J. Lightwave Technol. 36, 4843-4855 (2018) [https://www.osapublishing.org/jlt/abstract.cfm?URI=jlt-36-20-4843]

### Prerequisites

- Optical Communication Systems

## Lagrange Coded Computation (Lukas Holzbaur)

In distributed computing a user wishes to compute the evaluation of a function in a set of data points by employing multiple worker nodes. Several aspects related to this concept have been studied in recent literature, such as the resilience to stragglers, i.e., worker nodes that do not answer or have a large delay, and the privacy of the data, i.e., the evaluation points. One solution for the function class of polynomials was presented in [1], which can provide both resistance to stragglers and data privacy. The task of the student is to understand the properties and limitations of this framework.

### References

- [1] Yu, Qian, et al. "Lagrange coded computing: Optimal design for resiliency, security and privacy." arXiv preprint arXiv:1806.00939 (2018), arxiv.org/abs/1806.00939.

### Prerequisites

- Basic knowledge in algebra

## Multiple-Description Coding (Abdalla Ibrahim)

A class of distributed source coding problems. Consider the problem of generating two descriptions of a source such that each descriptionby itself can be used to reconstruct the source with some desired distortion andthe two descriptions together can be used to reconstruct the source with a lower distortion.This problem is motivated by the need to efficiently communicate multimedia content over networks such as the Internet. The problem has been examined in a series of papers since the 1970s, But it is still largely an open problem.

### References

- [1] Abbas El Gamal and Young-Han Kim, Lecture Notes on Network Information Theory (2010) , Chapter 14, and references at the end of the chapter arxiv.org/pdf/1001.3404v4.pdf [2] Abbas El Gamal and Young-Han Kim, Network Information Theory, Cambridge University Press (2011), Chapter 13, and Bibliographic notes at the end of the chapter. [3] Thomas Cover and Joy Thomas, Elements of Information Theory, 2nd ed., (2006), Chapter 10.

### Prerequisites

- Basic knowledge in Probability Theory and Information Theory.

## Identification Codes via Strongly Universal Hashing (Mohammad Javad Salariseddigh)

In contrast to Shannon Theory where code size grows only exponentially, in theory of identification, size of the code is a double exponential function in block length. This extra gain in performance of the code motivates studying of Identification theorems, codes and their explicit code schemes. Student will understand the basic concept of identification and common applications along with existing explicit code construction method based on idea of binary constant weight codes. epsilon-almost strongly universal classes of hash functions yield better explicit constructions of identification codes than well known version namely binary constant weight codes. Student should be able to understand how the algebraic definitions of ASU are and why it perform better in terms of code parameters.

### References

- [1] R. Ahlswede and G. Dueck, "Identification via channels," in IEEE Transactions on Information Theory , vol. 35, no. 1, pp. 15-29, Jan. 1989. (https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=42172) [2] S. Verdu and V. K. Wei, "Explicit construction of optimal constant-weight codes for identification via channels," in IEEE Transactions on Information Theory , vol. 39, no. 1, pp. 30-36, Jan. 1993.] (https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=179339) [3] K. Kurosawa and T. Yoshida, "Strongly universal hashing and identification codes via channels," in IEEE Transactions on Information Theory , vol. 45, no. 6, pp. 2091-2095, Sept. 1999. (https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=782144) [4] Identification via Channels and Constant-Weight Codes ( people.eecs.berkeley.edu/~ananth/229BSpr05/Reports/KrishEswaran.pdf)

### Prerequisites

- Basic Knowledge in Algebra and Channel Coding

## Rank-Based Encryption (Julian Renner)

Assuming an attack of a sufficiently large quantum computer, several classical public-key algorithms based on factoring large integers and the discrete logarithm problem become insecure since computationally intensive mathematical problems become easy-to-solve. Thus, the National Insitute of Standards and Technology (NIST) started the Post-Quantum Cryptography Standardization. The student should understand the cryptosystem based on LRPC codes [1] and determine its applications in the NIST proposals.

### References

- [1] N. Aragon, P. Gaborit, A. Hauteville, O. Ruatta and G. Zémor, "Low Rank Parity Check Codes: New Decoding Algorithms and Applications to Cryptography," in IEEE Transactions on Information Theory, vol. 65, no. 12, pp. 7697-7717, Dec. 2019 (https://ieeexplore.ieee.org/abstract/document/8793239)

### Prerequisites

- Basic Knowledge in Channel Coding

## Non-square Constellations for Adaptive Modulation (Delcho Donev)

Investigation of non-square constellations [1]. Bit-error rates for such constellations for transmissions over the AWGN channel.

### References

- [1] M. Abdelazis and T. A. Gulliver, "Triangular Constellations for Adaptive Modulation" ( ieeexplore.ieee.org/document/8067494)

### Prerequisites

- Digital Communications 1 and 2

## Turbo Codes for Insertion and Deletion Channels (Andreas Lenz)

Turbo codes [1,2] are a popular class of codes that achieve small error probabilities for rates close to the Shannon limit for, e.g., AWGN, BSC channels. On the other hand, coding for channels that introduce insertions and deletions remains poorly understood. The task of the student will be to understand standard turbo codes and review an approach for turbo decoding for insertion and deletion channels [3].

### References

- [1] Berrou, Claude, Alain Glavieux, and Punya Thitimajshima. "Near Shannon limit error-correcting coding and decoding: Turbo-codes. 1." Proceedings of ICC'93-IEEE International Conference on Communications. Vol. 2. IEEE, 1993. [2] Sklar, Bernard. "A primer on turbo code concepts." IEEE communications Magazine 35.12 (1997): 94-102. [3] Mansour, Mohamed F., and Ahmed H. Tewfik. "A turbo coding scheme for channels with synchronization errors." IEEE transactions on communications 60.8 (2012): 2091-2100.

### Prerequisites

- Channel Codng

## Interleaver Design of Turbo Codes (Peihong Yuan)

The LTE standard adopted a Turbo Code (TC) as channel code. And TCs could remain promising channel coding candidates for 5G. A typical TC is constructed by parallel concatenating two convolutional codes via an interleaver. The design of the interleaver is critical to the performance of the TC, particularly for short frame sizes. They can in general be separated into two classes: random interleavers and deterministic interleavers. The task of the student is to give an overview of three kinds of deterministic interleavers: Quadratic Polynomial Permutation (QPP) [1], Almost Regular Permutation (ARP) [2] and Dithered Relative Prime (DRP) lnterleavers [3].

### References

- [1] ieeexplore.ieee.org/document/1377495/ [2] ieeexplore.ieee.org/document/1312507/ [3] read.pudn.com/downloads155/doc/project/686322/DRP.pdf

### Prerequisites

- Channel Codes for Iterative Decoding

## Coded v.s. Uncoded Placement Schemes in Coded Caching (Hedongliang Liu)

The goal of coded caching is to minimize the transmission throughput during the peak hours while allowing users to pre-fetch some data during the idle hours. Using certain algebraic coding scheme in the placement phase of coded caching problems has been shown beneficial in terms of transmission throughput [3]. The task of this topic is to understand the coded placement schemes in [1] and [2] and give a description of the two schemes.

### References

- [1] J. Gómez-Vilardebó. Fundamental limits of caching: Improved rate- memory tradeoff with coded prefetching. IEEE Transactions on Com- munications , 66(10):4488–4497, Oct 2018. (https://ieeexplore-ieee-org.eaccess.ub.tum.de/abstract/document/8356039) [2 ] C. Tian and J. Chen. Caching and delivery via interference elimi-nation.IEEE Transactions on Information Theory, 64(3):1548–1560,March 2018. (https://ieeexplore-ieee-org.eaccess.ub.tum.de/document/8262655) [3] Q. Yu, M. A. Maddah-Ali, and A. S. Avestimehr. The exact rate-memory tradeoff for caching with uncoded prefetching.IEEE Trans-actions on Information Theory, 64(2):1281–1296, Feb 2018. (https://ieeexplore-ieee-org.eaccess.ub.tum.de/document/8226776)

### Prerequisites

- Linear Algebra, Channel Coding

## Computation of Information Rates by Particle Methods (Javier Garcia)

Particle filtering is an extension of trellis-based methods that allows computation of achievable rates of channels with memory and a continuous state space. The task of the student is to understand the principle of particle filtering, and describe the method and its applications.

### References

- [1] J. Dauwels and H. Loeliger, "Computation of Information Rates by Particle Methods," in IEEE Transactions on Information Theory, vol. 54, no. 1, pp. 406-409, Jan. 2008. ieeexplore.ieee.org/document/4418471 [2] P. M. Djuric et al., "Particle filtering," in IEEE Signal Processing Magazine, vol. 20, no. 5, pp. 19-38, Sept. 2003. ieeexplore.ieee.org/abstract/document/1236770

### Prerequisites

- Information theory

## Optical Phase Conjugation for Communications (Tasnad Kernetzky)

The student's first task is to get an understanding of the nonlinear effect called optical phase conjugation (OPC). Then, the goal is to present an overview of the principle and how it can be used to improve optical communications.

### References

- [1] Guang S. He - Optical phase conjugation: principles, techniques, and applications [2] B.N. Due - Nonlinearity compensation in DWDM metro systems using optical phase conjugation (https://ieeexplore.ieee.org/document/8924529) [3] Saber Rahbarfam - Nonlinear phase noise reduction using digital back propagation and midpoint optical phase conjugation (https://www.osapublishing.org/oe/abstract.cfm?uri=oe-27-6-8968) [4] T. Schneider - Nonlinear Optics in Telecommunications [5] G. P. Agrawal - Nonlinear Fiber Optics

### Prerequisites

- Course on Optical communications.